INDEX

Explanations

references to political entities or government-related words

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

acles

-0.96

furt

-0.81

borgh

-0.81

igmatic

-0.79

acity

-0.79

iating

-0.78

âĶģ

-0.78

ugu

-0.76

otten

-0.76

ativity

-0.75

POSITIVE LOGITS

wide

1.17

 sheriff

1.13

 commissioners

0.99

 clerk

0.97

 coroner

0.96

 Sheriff

0.94

 courthouse

0.94

 jails

0.94

 clerks

0.87

 jail

0.86

Activations Density 0.022%

references to political entities or government-related words

No Comments

No Known Activations